These types of questions come up often but due to the slight variances in everybody's situation I think I'll add to the pile.  Here goes:

My wife and I will be building a 2800ft² two-story house in eastern Ontario in about a year or so.  We have a reputable builder already lined up, but I'm trying to go as efficient as the budget allows.  I've read many of the building science articles on insulating a wall.  I was pretty set on the 2"+ exterior foam board (XPS or Polyiso) with sprayed-cellulose in the 2X6 cavities.  However, after discussing with my builder and some locals that went a "non-standard" route, it seems like the local interpretation of the building code requires a 6mil poly on the warm side of the insulation.  No ifs, ands or buts.  So, I think my initial plan kinda goes out the window since the above solution with an extra vapor barrier on the inside doesn't permit the system to dry to the inside or outside (even more of a problem with wet cellulose). 

Now, I didn't look at all the numbers but would the following variation on the original plan work?  1.5" XPS on the outside with OSB sheathing, 2X6" studs with 5.5" of Roxul rock wool and the poly.  This gives a cavity insulation of about R-30 (anybody know the whole wall value)?  Would this allow the system to dry towards the outside?  Any other solution (keeping in mind that the budget isn't unlimited)?

Thanks,

-Mike

first of all...you're in the right forum. there are few people around here that can help you.

I wouldn't build something that can't dry in at least one direction. if you think that there is no way you can convince your building inspectors to let you skip the vapour barrier
(I've seen projects in toronto where they asked poly over 5.5" of closed cell foam) than you can maybe modify a bit the exterior layers.

how about use some roxul boards an top of woodboard sheating instead of 1.5" XPS on the outside with OSB sheathing. same R value but much better permeability

http://bpcan.com/en-CA/products/insulation-and-structural-boards/sheathing-and-panels/
http://www.roxul.com/building+envelope/products/roxul+cavityrock%C2%AE

both products are manufactured in ontario or quebec and you should be able to stay in the same price range.

good luck,
Adrian

It's the first I hear of Roxul CavityRock, but it seems the documentation leads to believe that it's intent is mostly for insulating over metal studs. Regardless of this, would it work to allow interior moisture to escape outwards and "repel" water vapour that could be driven through the exterior cladding?

As for the BP sheathing, how does it compare in price to a similar size sheath of OSB?

What's your exact location or nearest town (for weather data purposes.) At 2" of iso you might be just on the edge of being able to make the average dew-point argument for a class-III vapor retarder, and surely you'd be able to get by with Certainteed Membrain or a vapor-barrier latex almost anywhere in eastern Ontario (and standard latex at any latitude south of Toronto.)

As I understand it the national building code (still) calls out 6-mil poly as an interior vapor barrier everywhere in Canada, despite strong evidence that it's not necessary and even counterproductive in places like Vancouver. SFAIK the IRC 2009 has yet to be fully integrated into the national codes. But vapor barrier latex (~0.5 US perms) should be good enough. Unfortunately the letter of the law spells out LOWER vapor retardency when the exterior is relatively low-permeance (such as foil-facers or 2" of XPS), in contravention with good building-science:

----------------------------- begin clipped text------------------

Section 5.5 Vapour Diffusion

5.5.1 Vapour Barriers

5.5.1.1. Required Vapour Barrier

1) Except as provided in Sentence (2), where a building component or assembly will be subjected to a temperature differential and a differential in water vapour pressure, the component or assembly shall include a vapour barrier.

2) A vapour barrier is not required where it can be shown that controlled vapour diffusuion will not adversely affect any of

a) the health or safety of the building users,
b) the intended use of the building , or
c) the operation of building services.

5.5.1.2. Vapour Barrier Properties and Installation
(See A-5.3.1.2 in Appendix A)

1) The vapour barrier shall have sufficiently low permeance and shall be positioned in the building component or assembly so as to

a) minimize moisture transfer by diffusion, to surfaces within the assembly that would be cold enough to cause condensation at the design temperature and humidity conditions, or
b) reduce moisture transfer by diffusion, to surfaces within the assembly that would be cold enough to cause condensation at the design temperature and humidity conditions, to a rate that would not allow siffucient accumulation of moisture to cause deterioration or otherwise adversely affect any of

i) the health or safety of the building users
ii) the intended use of the building , or
iii) the operation of building services.
(See Appendix A.)

9.25.4 Vapour Barriers

9.25.4.1. Required Barrier to Vapour Diffusion

1) Thermally insulated wall, ceiling and floor assemblies shall be constructed with a vapour barrier so as to provide a barrier to diffusion of water vapour from the interior into wall spaces, floor spaces or attic or roof spaces.

9.25.4.2. Vapour Barrier Materials

1) Except as required in Sentence (2), vapour barriers shall have an initial permeance not greater than 45 ng/(Pa-s-m 2 ).

2) When used where a high resistance to vapour movement is required, such as in wall constructions that incorporate exterior cladding or sheathing having a low water vapour permeance, vapour barriers shall have a permeance not greater than 15 ng/(Pa-s-m 2 ). (See Appendix A.)

9.25.4.3. Installation of Vapour Barriers

1) Vapour barriers shall be installed to protect the entire surfaces of thermally insulated wall, ceiling and floor assemblies.

2) Vapour barriers shall be installed sufficiently close to the warm side of insulation to prevent condensation at design conditions. (See Apendix A.)

-----------------------------end clipped text------------------

Section 9.25.4.3 could be an out, were it not for the fact that it calls for the prevention of condensation at DESIGN conditions, rather than at AVERAGE MIDWINTER conditions. The heating design temp is considerably colder than the average January-February temp, but it's really the average over the 6-10 coldest that counts in terms of raising risk of rot mold or moisture damage. And with cellulose cavity fill any condensation that DOES occur wicks away and is redistributed without damage to the wood or the insulation itself. If it spelled out the climate-average rather than design temp you may already have a sufficient R-foam/R-cellulose ratio. The code is too severely drawn.

If you can't get the inspectors to sign off on a class-II vapor retarder such as barrier-latex, use 3"/76mm of unfaced EPS on the exterior to achieve similar thermal performance that lets it dry toward the exterior. With the 10mm code-min rainscreen the cellulose would still dry through that much EPS (even 100mm would be OK, but not much more than that.)

Posted By Dana1 on 05 Oct 2011 11:16 AM
What's your exact location or nearest town (for weather data purposes.) At 2" of iso you might be just on the edge of being able to make the average dew-point argument for a class-III vapor retarder, and surely you'd be able to get by with Certainteed Membrain or a vapor-barrier latex almost anywhere in eastern Ontario (and standard latex at any latitude south of Toronto.)

I'm about 50kms east of Ottawa.

Just a question that my builder brought up when discussing foam board thicknesses. How wide are typical foundations on these builds with 2.5+" foam boards (assuming stone veneer cladding)?

The quick & dirty (and mostly correct) calc:

The historical mean temp for the month of January in Ottawa is about -14C/(+7F), so with R10 of foam outside of R20 of cellulose, with a mean interior temp of 20C the mean temp at the sheathing will be about -3C. Assuming 35% interior relative humidity the dew point of 20C 35% RH air is about +4C (~39F), so 2 inches of XPS isn't going to cut it for being able to run with plain-old-latex on the interior.

Bump that to R20 (4") and you'd be in the right range (close enough to not worry about it, given the moisture buffering of the cellulose.) With 4" of polyiso (R22.5, when derated for cold weather temp) you'd be fine, but still expect an argument from the inspector.

Some will even use the mean temp for the coldest 10 or 12 weeks of winter, which yields results similar to IRC prescriptions for US climate zones. This is the crudest possible model, and not an accurate representation of what's really going on, but it's actually remarkably good at delivering a correct result (within a couple of percent) without a lot of work. I prefer to stick with the January mean-temp dewpoint calc, since it gives a bit more margin than IRC minimums or calculating it based on the 3-month mean temp. In Ottawa the 3-month mean is about -10C which would bring the mean temp at the sheathing only up to 0C with 2" of exterior XPS.

It really needs more than R10 to work without at least a class-II vapor retarder. But at R10 it's close enough that you really don't need a class-I retarder like poly. You could go nuts on it, run a WUFI simulation (or pay for an engineer to run the sim) to prove it, but it's doubtful than an inspector would accept that in the face of the letter-of-the law. Design temps are much lower than the -14C mean.

Why dont you just pull the poly after your insulation inspection clears?

Looking into various products now and the Dow Cladmate XPS foam board is rated at 3.5 perm @ 1". Is there a formula that would tell us it's permeance at various other thicknesses (ie: 1.5", 2")? I'm not willing to go any thicker than 3"... mostly due to costs associated with going thicker than that (wider foundation, longer screws, etc).

It's roughly linear- doubling the thickness cuts the permeance roughly in half (and conversely.) At 3" you'd still be at or above 1 perm.

At 1 perm things can still dry quite reasonably, but isn't so vapor-open that large amounts of moisture can pass in or out in the course of a day, but over a few weeks quite a bit can when the difference in vapor pressure is high. Over seasonal averages it's pretty good, but it's tight enough to reduce a few week's worth of peak diffusion. A few few hours of sun-baking on rain-soaked fiber cement siding passes very little moisture through 1-perm or even 1.75 perm vapor retarders to condense on the air-conditioned side of the assembly, whereas with a ~10perm housewrap like Typar (or 40 perm Tyvek) it's an issue. Traditional 15# felt underlayment is about 5 perms. OSB sheathing is about 2-5 perms all by itself, depending on it's moisture content.

By going as low as 1 perm you've reduced the moisture transport by ~80%, but that's still reasonable drying rate in seasonal terms. At 0.1 perms the amount of moisture that would have dried in one week at a given difference in vapor pressure with a 1-perm retarder now takes 10 weeks.

Polyethylene sheeting at 6 mils thickness is about 0.06 to 0.08 perms, so now you're talking more than 3 months to move what a 1-perm retarder would allow in one week. Sure it blocks moisture from getting in at a rapid rate, but it's too severe. In the vast majority of houses with interior poly the poly has (or will have) pinhole leaks, nicks and tears, and the air-transported moisture from even those small leaks will be orders of magnitude higher than the vapor diffusion through the poly when the humidity of the air is high. But with poly being so vapor-tight, it won't let that moisture back out nearly as quickly when conditions change.

With half-perm paint on the interior and 1-perm insulating sheathing on the exterior there's something of a flow-through aspect happening in winter, but it'll still dry in both directions come warmer weather. Raising the permeance of the sheathing to 1.75 perms (or 2) won't make a dramatic performance difference.